Quick viewing(Text Mode)

An Easy-To-Use Genotoxicity Assay Using EGFP-MDC1-Expressing Human Cells

An Easy-To-Use Genotoxicity Assay Using EGFP-MDC1-Expressing Human Cells

and Environment, Vol. 36, No. 1 pp. 17–28 (2014)

Regular article An Easy-to-use Assay Using EGFP-MDC1-expressing Human Cells

Shun Matsuda1,RyoMatsuda2, Yoko Matsuda1, Shin-ya Yanagisawa1, Masae Ikura2, Tsuyoshi Ikura2 and Tomonari Matsuda1,3 1Research Center for Environmental Quality Management, Kyoto University, Otsu, Shiga, Japan 2Radiation Biology Center, Kyoto University, Kyoto, Japan

Received December 26, 2013; Revised January 16, 2014; Accepted January 21, 2014 J-STAGE Advance published date: February 1, 2014

Histone H2AX phosphorylated at Ser139 (g-H2AX) is a gregates called foci (1,2). Because a single g-H2AX useful biomarker for DNA double-strand breaks. However, focus corresponds to one DSB (2,3), g-H2AX is an ex- g-H2AX detection has methodological disadvantages such tremely sensitive indicator for DSBs. Whereas induction as the requirement of expensive anti-g-H2AX antibody and of g-H2AX was ˆrst shown in response to ionizing time-consuming handling for its staining. Mediator of DNA irradiation (1), it has since been reported that a broad damage checkpoint 1 (MDC1) is a central adaptor spectrum of mutagens also induces g-H2AX, including which recruits various DNA damage response to radical generators [bleomycin (4), tirapazamine (4), g-H2AX and thus forms nuclear foci in the same location as g-H2AX in response to DNA damage. Here, we describe calicheamicin g1 (5), c-1027 (6), and neocarzinostatin an easy-to-use genotoxicity assay which combines en- (7)], topoisomerase inhibitors [camptothecin (CPT) (6), hanced green ‰uorescence protein (EGFP)-fused MDC1- topotecan (8), and etoposide (4)], a DNA intercalator expressing cells with a free R program for image-proc- [ (9)], alkylating agents [N-methyl-N- essing and quantiˆcation of foci area/nucleus. The wor- nitrosourea (MNU) (10), N-methyl-N?-nitro-N- k‰ow of this assay is simple: mutagen treatment, imaging, nitrosoguanidine (11), adzelesin (6), methyl methan- and R-processing. This assay does not need antibodies or esulfonate (12), and N-ethyl-N-nitrosourea (13)], an staining handling and it detected the genotoxicity of a oxidizing agent [hydrogen peroxide (7)], bulky DNA ad- range of mutagens, including camptothecin (topoiso- duct-forming agents [4-nitroquinoline1-oxide (4NQO) merase I inhibitior), (crosslinker), and 4-nitro- (4), benzo[a]pyrene (B[a]P) (13), and N-acetoxy-2- quinoline 1-oxide and benzo[a]pyrene (bulky DNA-adduct acetylamino‰uorene (13)], DNA crosslinking agents forming compounds), as increased ‰uorescence of EGFP- MDC1 foci. Furthermore, cotreatment with arabinofu- [cisplatin (CDDP) (14) and actinomycin D (15)], a ranosyl cytosine/hydroxyurea and mutagens sensitized ribonucleotide reductase inhibitor [hydroxyurea (HU) EGFP-MDC1 foci formation to bulky DNA adduct-type (16)], a DNA polymerase inhibitor [aphidicolin (17)], mutagens. Additionally, the established cells can be moni- and a general tyrosine inhibitor [staurosporine tored in real-time using live cell imaging to obtain detailed (18)]. Furthermore, the relationship between chemical dynamics of MDC1 in response to mutagens. The simple dose and g-H2AX induction is linear (4). These facts handling of this assay is expected to enable its full automa- explain why there is increasing interest in using g-H2AX tion, thus making it useful for high-throughput genotoxici- as a biomarker in chemical genotoxicity assays. Indeed, ty screening of chemicals and monitoring of environmental previous studies have used g-H2AX in genotoxicity as- mutagens. says for photogenotoxic chemicals (19,20) and for known and potential environmental pollutants such as Key words: MDC1, g-H2AX, ‰uorescence microscopy, benzene (21), polyaromatic hydrocarbons (22), nonyl- image-processing, R phenol polyethoxylates (23), heavy metals (24), cigarette smoke (25), and nanoparticles (26–32), and the results

were all positive except for alumina (Al2O3) nanoparti- Introduction cles. H2AX phosphorylated at Ser139 (g-H2AX) 3 is one of the most useful biomarkers for DNA double- Correspondence to: Tomonari Matsuda, Research Center for En- vironmental Quality Management, Kyoto University, Otsu, Shiga strand breaks (DSBs). When DSBs are induced in cells, 5200811, Japan. Tel: ＋81-77-527-6224, Fax: ＋81-77-524-9869, H2AX polypeptides within about ¿2 Mbp of a DSB E-mail: matsuda.tomonari.8z＠kyoto-u.ac.jp site are phosphorylated to form g-H2AX nuclear ag- doi: org/10.3123/jemsge.2014.001

 The Japanese Environmental Mutagen Society 17 Shun Matsuda et al.

On the other hand, genotoxicity assays targeting g- the tandem BRCA1 C-terminus (BRCT) domains inter- H2AX still have several methodological limitations. act with g-H2AX (39,40), (54), p53-binding protein Immuno‰uorescence microscopy, ‰ow cytometry, 1 (55), cell division cycle 27 (56), murine double minute western blotting, and enzyme-linked immunosorbent 2 (57), and Ser 1524-phosphorylated topoisomerase IIa assay (ELISA) are conventionally used for g-H2AX de- (58). tection, and all these methods absolutely require anti-g- Here, we established human cells expressing enhanced H2AX antibody, which is an expensive consumable. green ‰uorescent protein (EGFP)-fused MDC1 (EGFP- Furthermore, staining with g-H2AX antibody is also MDC1) and developed an easy-to-use and rapid geno- burdensome and time consuming and high-throughput toxicity assay, which combines both the cells and a free screening with these methods would be di‹cult. For en- R program for imaging and statistical analysis (59). We vironmental and chemical management applications, an then demonstrated the e‹cacy of the genotoxicity assay easy-to-use and high-throughput method for chemical by testing several diŠerent types of mutagens, including genotoxicity screening would be advantageous. the DNA topoisomerase inhibitor CPT, the crosslinking Mediator of DNA damage checkpoint 1 (MDC1) is a agent CDDP, the bulky adduct-forming compounds large protein (2,089 amino acids) (33) which plays a 4NQO and B[a]P, and the alkylating agent MNU. central role in ampliˆcation and mediation of the DNA damage response (DDR). In human cells, when a DSB Materials and Methods occurs, the MRN complex composed of MRE11 (meiot- Chemicals: Dimethyl sulfoxide (DMSO), MNU, ic recombination 11)/RAD50 (homolog of S. arabinofuranosyl cytosine (AraC) and HU were pur- cerevisiae's Rad50)/NBS1 (Nijmegen breakage syn- chased from Wako (Osaka, Japan). CDDP and CPT drome protein 1) binds to the DSB site (34), whereupon were purchased from Sigma-Aldrich (Missouri, USA). ataxia telangiectasia mutated (ATM) protein is recruited 4NQO and B[a]P were purchased from Nacalai Tesque to the DNA lesion via NBS1 (35,36). ATM phosphory- (Kyoto, Japan). O6-benzylguanine (O6-BG) was pur- lates histone H2AX adjacent to the DSB site to form g- chased from Santa Cruz (California, USA). The chemi- H2AX (37). At the same time, DNA-dependent protein cals were dissolved in DMSO. kinase (DNA-PK) is also recruited to the DSB site and Cell culture: Human breast adenocarcinoma MCF7 phosphorylates H2AX (38). Subsequently, MDC1 binds cells were cultured in Dulbecco's Modiˆed Eagle's to g-H2AX (39) and brings additional ATM to the DSB Medium supplemented with 10z fetal bovine serum at site, expanding ATM-dependent H2AX phosphoryla- 379C in a humidiˆed 5z CO2 incubator. tion (40). The positive feedback loop involving ATM-g- Construction of pEGFP-C1/MDC1 plasmid: First, H2AX-MDC1 ampliˆes DDR signaling. In the case of the middle region of human MDC1 cDNA (KIAA0170) replication stress, ATM and rad3-related (ATR) protein were cut out with HindIII and EcoRIandinsertedinto phosphorylates H2AX at stalled replication forks the same sites of pEGFP-C1 vector (Clontech, Califor- (9,16). A recent report demonstrated that MDC1 binds nia, USA) (pEGFP-C1/MDC1 DNDC). Next, the 3? to topoisomerase II binding protein 1 (TopBP1) and terminus region of human MDC1 was ampliˆed from that both H2AX and MDC1 are required for formation human MDC1 cDNA by polymerase chain reaction of TopBP1 foci at stalled replication forks (41), suggest- (PCR) with the following primers: 3? terminus MDC1- ing a role for MDC1 in response to replication stress. By F, 5?-ccggaattccaatctcctgtcaccacagaccag-3? (an EcoRI these mechanisms, MDC1 immediately (within 5 min site is underlined); 3? terminus MDC1-R, 5?-tccccgcgg after ionizing irradiation) forms foci in the same loca- tcaggtggatgacatctccaaagggg-3? (a SacII site is under- tions as those formed by g-H2AX in response to DSBs lined). The PCR product was cut with EcoRI and SacII (2,42). andinsertedintopEGFP-C1/MDC1DNDCcutwith MDC1 has multiple domains for mediating pro- the same restriction enzymes (pEGFP-C1/MDC1 DN). tein–protein interactions and thus functions as an im- Subsequently, the 5? terminus region of human MDC1 portant adaptor which brings and retains various DDR was ampliˆed from human MDC1 cDNA by PCR with proteins to DSB sites through g-H2AX. The N-terminal the following primers: 5? terminus MDC1-F, 5?-ccc forkhead-associated domain interacts with Ser1981- aagcttccgaggacacccaggctattgactgg-3? (a HindIII site is phosphorylated ATM (40,43) and Thr68-phosphorylat- underlined); 5? terminus MDC1-R, 5?-cccaagctt ed checkpoint kinase 2 (Chk2) (44) in response to DSBs ggcttttctccagagggacagcc-3? (a HindIII site is under- and constitutively with RAD51 recombinase (45); the lined). The PCR product was cut with HindIII and in- Ser-Asp-Thr repeats constitutively phosphorylated by serted into pEGFP-C1/MDC1 DN cut with the same casein kinase 2 interact with NBS1 (46–49); the Thr-Gln- restriction enzyme (pEGFP-C1/MDC1). Xaa-Phe repeats phosphorylated by ATM in response to Transduction of stable EGFP-MDC1-expressing DSBs interact with ring ˆnger protein 8 (50–52); the MCF7 (EGFP-MDC1/MCF7) cells: pEGFP-C1/ Pro/Ser/Thr repeats interact with DNA-PK (53); and MDC1 was transfected into MCF7 cells by Lipofecta-

1818 Genotoxicity Assay Using EGFP-MDC1 mine 2000 (Life technologies, California, USA) accord- LAS-1000 image analyzer controlled by Image Reader ing to the manufacturer's protocol. After 48 h, stable LAS-1000 Mini V1.21 software (Fujiˆlm, Tokyo, transformants were selected using medium supplement- Japan). ed with 1 mg/ml G418. Immuno‰uorescence staining and microscopy: Immunoprecipitation: Cellsgrownto80z con- For immuno‰uorescence staining, EGFP-MDC1-ex- ‰uence in 100 mm dishes were suspended into 1 mL ice- pressing MCF7 cells (2×105) were seeded onto a 24×24 cold phosphate-buŠered saline (PBS) (137 mM NaCl, mm2 cover glass (Matsunami glass, Osaka, Japan) in

2.68 mM KCl, 8.10 mM Na2HPO4,and1.67mM each well of a 6-well plate and treated with 10 mMCPT NaH2PO4) and pelleted by centrifugation at 1,000 rpm for 1 h. CPT dissolved in DMSO were added to the for 1 min at 49C. Cell pellets were resuspended in 1 mL medium directly and the ˆnal DMSO concentration was ice-cold cell lysis buŠer (150 mM NaCl, 10 mM Tris- 0.1z (v/v). After experimental treatment, subsequent HCl pH 7.3, 0.5 mM ethylenediaminetetraacetic acid, 1 incubation was performed in the dark as much as possi- mM beta-mercaptoethanol, 0.2 mM phenylmethylsul- ble. Cells were ˆxed with 4z formaldehyde in PBS for fonyl ‰uoride, 10 mM beta-glycerophosphate, and 1z 10 min and permeabilized with 0.2z Triton X-100 in NP-40) and incubated for 20 min on ice. Lysate was PBS for 10 min. Next, cells were incubated in 4z bo- centrifuged at 15,000 rpm for 15 min at 49C. 65 mLof vine serum albumin in PBS for 30 min for blocking. the supernatant was saved as a whole-cell lysate (Input) Cells were incubated for 60 min with anti-MDC1 an- sample and the remaining supernatant was used for im- tibody (1:100) (ab11169) or anti-g-H2AX antibody munoprecipitation. 1 mL anti-GFP antibody (ab290, (1:100) (a component of the OxiSelect DNA Double Abcam, Cambridge, UK) was added to the supernatant Strand Break (DSB) Staining Kit, Cell Biolabs, Califor- and the sample was rotated for 60 min at 49C. Next, af- nia, USA) in 1z BSA in PBS. As a negative control, an ter addition of 20 mLofa50z slurry of Protein A equal concentration of normal rabbit IgG (sc-2027, San- agarose (Merck Millipore, Massachusetts, USA), the ta Cruz) or normal mouse IgG (sc-2025, Santa Cruz) sample was rotated for 60 min at 49C.Thesamplewas diluted with 1z BSA in PBS was used. Subsequently, pelleted by centrifugation at 8,000 rpm for 1 min at cells were incubated for 60 min with tetramethyl- 49C. Pellets were washed with 500 mL ice-cold cell lysis rhodamine-5-(and 6)-isothiocyanate (TRITC)-conjugat- buŠer three times and then analyzed by Western blot- ed goat anti-rabbit (1:200) (ab6718, Abcam) or TRITC- ting as immunoprecipitated (IP) samples. conjugated goat anti-mouse IgG (ab7065, Abcam). Western blotting: Input or IP sample was dena- Nucleiwerestainedfor10minwith0.5mg/mL Hoechst tured in NuPAGE LDS Sample BuŠer (Life technol- 33342. Microscopic observations were performed using ogies) at 709C for 10 min. For IP samples, each sample a BZ-9000 ‰uorescence microscope (Keyence, Osaka, was pelleted by centrifugation at 15,000 rpm for 1 min Japan). A 20× (Plan Apo/0.75 numerical aperture) ob- and the supernatant was diluted ten-fold with NuPAGE jective was used. Image processing was performed by LDS Sample BuŠer. To separate proteins, 3 mL/sample using a BZ-II Analyzer (Keyence). wasappliedtoeachwellofaNuPAGENovex4–12z Programming in R: After microscopic imaging, Bis-Tris Gel (Life technologies) and electrophoresis was TIF-images were converted to JPG-images using BZ-II performed in NuPAGE MES SDS Running BuŠer (Life Resizer (Keyence) to reduce ˆle sizes. Image-processing technologies) at 200 V for 60 min. Separated proteins and statistical analysis were processed in R,andimage- were blotted onto Hybond ECL membrane (GE processing was performed using the EBImage package Healthcare, Buckinghamshire, UK) by electrophoresis (60). JPG-images for each individual treatment were (100 V for 90 min in blotting buŠer (20z methanol, 25 separately placed into diŠerent R-work folders. Wor- mM Tris, and 192 mM glycine)). The membrane was in- k‰ow of the program using R isshowninFig.2A.The cubated for 60 min with 5z skimmed milk in TBS-T detailed program was described in supplemental infor- (137 mM NaCl, 20 mM Tris-HCl pH 7.6, 0.1z Tween mation (Available at https://www.jstage.jst.go.jp/ 20) for blocking. Next, the membrane was incubated for browse/jemsge). Brie‰y, green channels of the EGFP- 60 min with anti-MDC1 antibody (1:1,000) (ab11169, MDC1 images were converted to gray scale images, Abcam) or anti-GFP antibody (1:1,000) (ab290, Ab- brightness of the images was normalized, and nuclei and cam) in TBS-T. Subsequently, the membrane was incu- foci were separately extracted from the same normalized bated for 60 min with horseradish peroxidase-conjugat- images. In nuclear extraction, the images were binarized ed goat anti-rabbit antibody (1:10,000) (#7074, Cell Sig- and the resulting sets of combined pixels (named ob- naling Technology, Massachusetts, USA) in TBS-T. To jects) were numbered. Based on thresholds of area and develop chemiluminescence by horseradish peroxidase, acircularity of each object and mean brightness of each the membrane was incubated for 5 min with ECL Prime nucleus corresponding to each object, noise (objects not WB Detection Reagent (GE Healthcare). corresponding to any nuclei in the normalized images) Chemiluminescence detection was performed using an and objects corresponding to nuclei with weak bright-

1919 Shun Matsuda et al. ness were removed. In foci extraction, the normalized performed using R. images were binarized using the diŠerent condition from that of nuclear extraction. Finally, total foci Results area/nucleus was calculated using the extracted nuclei To establish EGFP-MDC1/MCF7 cells, pEGFP- and foci images. C1/MDC1 was transfected into MCF7 cells. Although A new genotoxicity assay: EGFP-MDC1/MCF7 stable transformants had been selected by G418, the cells (2×105) seeded in 35-mm glass-base dishes were population of EGFP-MDC1-expressing cells observed treated with mutagens for various times. The mutagens by ‰uorescence microscopy was very small (about 1z), dissolved in DMSO were added to the medium directly but cells with a strong EGFP signal were isolated. To and the maximum of ˆnal DMSO concentration was conˆrm that full-length EGFP-MDC1 was expressed in 0.25z (v/v). In the AraC/HU co-treatment experi- the isolated cells, western blotting (WB) following im- ment, the ˆnal DMSO concentration was 0.4z (v/v). In munoprecipitation (IP) with anti-GFP antibody was the O6-BG co-treatment experiment, the ˆnal DMSO performed (Fig. 1A). In the input sample, anti-MDC1 concentration was 0.35z (v/v). Cells were ˆxed with antibody detected endogenous MDC1 signal located at 4z formaldehyde in PBS for 10 min. After a PBS wash, about 220 kDa in both EGFP-MDC1-transduced cells microscopic observations were performed. A 20× (Plan and negative control cells, whereas an additional MDC1 Apo/0.75 numerical aperture) objective and the quick signal corresponding to a protein larger than en- full-focus function were used. In each experimental con- dogenous MDC1 was detected only in EGFP-MDC1- dition, 10–15 diŠerent ˆelds were imaged and image- transduced cells. The diŠerence in sizes between the en- processing and calculation of foci area/nucleus were dogenous and the larger MDC1 corresponds to the size

Fig. 1. (Color online) Establishment of EGFP-MDC1-expressing MCF7 cells. (A) Expression of EGFP-MDC1 in human breast adenocarcinoma MCF7 cells. Cells were transduced with pEGFP-C1/MDC1 plasmids. After G418 selection, the cells were lysed, immunoprecipitated with anti- GFP antibody, and the precipitates were subjected to western blotting using anti-GFP or anti-MDC1 antibodies. The opened star indicates bands corresponding to endogenous MDC1. (B) EGFP signal forms foci in response to DNA damage and colocalizes with MDC1. After treatment for 1 h with 0.1z (v/v) DMSO (vehicle control) or 10 mM camptothecin (CPT), EGFP-MDC1/MCF7 cells were immunostained with anti-MDC1 an- tibody. (C) EGFP signal colocalizes with g-H2AX. After treatment for 1 h with 0.1z (v/v) DMSO (vehicle control) or 10 mM CPT, EGFP- MDC1/MCF7 cells were immunostained with anti-g-H2AX antibody.

20 Genotoxicity Assay Using EGFP-MDC1

Fig. 2. (Color online) Work‰ow of image-processing using R. (A) Work‰ow of the program using R. After EGFP-MDC1 imaging, nuclei and foci were separately masked from the same EGFP-MDC1 image and then the masked foci area in each masked nucleus was calculated. (B) Representative images of masked nuclei and masked foci of normal nuclei and nuclei forming foci. of EGFP (about 27 kDa). Using anti-GFP antibody, As shown in Fig. 1B, immunostaining with anti-MDC1 GFP signal was detected only in EGFP-MDC1-tran- antibody produced a diŠuse nuclear signal in DMSO- sduced cells, and the band was located at the same size treated cells but revealed discrete foci in nuclei in as the upper MDC1-signal. After IP with anti-GFP an- response to CPT. Localization of GFP signal complete- tibody, both GFP and MDC1 signals were detected only ly overlapped with that of the anti-MDC1 signal. Simi- in EGFP-MDC1-transduced cells and their locations larly, the signal from anti-g-H2AX antibody was ob- corresponded to that of the larger MDC1. These results served as discrete foci and increased in response to CPT clearly show that full-length EGFP-MDC1 was ex- (Fig. 1C). Foci identiˆed by GFP signal overlapped with pressedinEGFP-MDC1-transducedcellsandthelevel the anti-g-H2AX signal. Given these facts, we con- of EGFP-MDC1 expression was much higher than that ˆrmed that EGFP-MDC1 expressed in EGFP- of endogenous MDC1. MDC1/MCF7 cells was functional, at least to the extent We tested whether EGFP-MDC1 expressed in the es- that EGFP-MDC1 localizes to nuclear foci in response tablished EGFP-MDC1/MCF7 cells was functional. to DNA damage. Cells were treated with DMSO (negative control) or 10 Next, to quantify EGFP-MDC1 formation after mM CPT for 1 h and immuno‰uorescence analysis using mutagen treatment, we developed an automated pro- anti-MDC1 and anti-g-H2AX antibody was performed. gram for image-processing using R. The outline of the

21 Shun Matsuda et al.

Fig. 3. Results of mutagen tests using the new genotoxicity assay. (A) Representative histograms of EGFP-MDC1 foci area/nucleus of EGFP- MDC1/MCF7 cells treated for 1 h with 0.1z (v/v) DMSO (control) or 10 mM CPT (damage). (B) Boxplots of foci area/nucleus for each condi- tion. After EGFP-MDC1/MCF7 cells were treated for 1, 3, or 24 h with 0.25z (v/v) DMSO (vehicle control), 10 mM 4NQO, 10 mM CPT, 40 mM CDDP, or 1 mM MNU, the new genotoxicity assay was performed. 200–490 cells were counted for each condition. (C) Boxplots of foci area/nucleus of EGFP-MDC1/MCF7 cells treated with B[a]P. EGFP-MDC1/MCF7 cells were either left untreated (0 h) or treated for 1, 3, or 24 h with 1 mMB[a]P, and the new genotoxicity assay was performed. 271–432 cells were counted for each condition. *pº0.0001 versus 0 h, **pº1.0 ×10－12 versus DMSO, Wilcoxon signed-rank test, respectively. procedure is shown in Fig. 2A, and works as follows. gens: CPT (topoisomerase I inhibition), 4NQO (bulky After mutagen treatment, GFP signal in the cells is im- DNA-adduct formation), CDDP (DNA crosslinking), aged by immuno‰uorescence microscopy with vivid and MNU (DNA alkylation). EGFP-MDC1/MCF7 cells clarity. Then, each nucleus area and focus area are weretreatedwith10mMCPT,10mM 4NQO, 40 mM separately extracted from the same image (Fig. 2B), CDDP, or 1 mM MNU for 1, 3, or 24 h, and the geno- each focus is linked to its corresponding nucleus, and toxicity assay was performed. Representative and all total foci area in each nucleus is calculated using R.The histograms of EGFP-MDC1 foci area/nucleus of nor- details of the procedure are described in Materials and malanddamagedcellsareshowninFig.3Aand Methods. Furthermore, we carefully optimized the con- Supplemental Fig. 1, clearly illustrating that most dition of the image-processing conditions of the pro- DMSO-treated (control) cells have relatively small gram. EGFP-MDC1 foci areas whereas most CPT-treated Using EGFP-MDC1/MCF7 cells and the image-proc- (damaged) cells have extensive foci areas. As shown in essing program, we established a new genotoxicity as- Fig. 3B, CPT or 4NQO exposure signiˆcantly increased say. The work‰ow is simple: sample treatment, ˆxation the area of EGFP-MDC1 foci after 1 h whereas no obvi- (if required), microscopic imaging, and image-proc- ous changes were detected after CDDP or MNU treat- essing and calculation of foci area/nucleus using R.We ment at 1 h. However, CDDP treatment signiˆcantly in- applied the genotoxicity assay to four types of muta- creased the area of EGFP-MDC1 foci after 24 h. On the

22 Genotoxicity Assay Using EGFP-MDC1

Fig. 4. Enhancement of EGFP-MDC1 foci formation. (A) Boxplots of foci area/nucleus of EGFP-MDC1/MCF7 cells cotreated with 4NQO and AraC/HU. After EGFP-MDC1/MCF7 cells were cotreated for 3 h with 10 mMAraC/2mMHUand0(0.1z (v/v) DMSO, vehicle control), 0.1 or 1 mM 4NQO, the new genotoxicity assay was performed. 783–1069 cells were counted for each condition. (B) Boxplots of foci area/nucleus of EGFP-MDC1/MCF7 cells cotreated with B[a]P and AraC/HU. EGFP-MDC1/MCF7 cells were cotreated for 3 h with 10 mMAraC/2mMHU and 0.1z (v/v) DMSO (vehicle control) or 10 mMB[a]P, and the new genotoxicity assay was performed. 189–204 cells were counted for each con- dition. (C) Boxplots of foci area/nucleus of EGFP-MDC1/MCF7 cells cotreated with MNU and O6-BG. EGFP-MDC1/MCF7 cells were untreated (Control)orcotreatedfor1,3,or24hwith25mM O6-BG and 0.25z (v/v) DMSO (vehicle control) (D) or 1 mM MNU (M), and the new genotox- icity assay was performed. 290–633 cells were counted for each condition. *pº0.05, **pº1.0×10－12 versus DMSO, #pº1.0×10－12 versus the same concentration of 4NQO without AraC/HU, Wilcoxon signed-rank test, respectively.

other hand, MNU did not increase the area of EGFP- niˆcantly increased EGFP-MDC1 formation after only MDC1 foci after 24 h, the longest time point examined. 3 h treatment (Fig. 4B). In an attempt to enhance the Next, we tested B[a]P which requires metabolic activa- sensitivity of EGFP-MDC1/MCF7 cells to MNU, which tion to form bulky DNA adducts. Whereas no increase was negative in the genotoxicity assay (Fig. 3B), 25 mM in the area of EGFP-MDC1 foci was observed in 1 mM O6-BG was co-administered with MNU. MNU forms B[a]P-treated cells at 1 h and 3 h, a signiˆcant increase alkylated DNA lesions such as O6-methylguanine (O6- was observed 24 h after treatment (Fig. 3C). MG) and O6-BG is an inhibitor of O6-methylguanine To enhance the sensitivity of EGFP-MDC1/MCF7 DNA methyltransferase (MGMT) which repairs O6-MG cells to mutagens, AraC (DNA polymerase inhibitor) back to guanine. As shown in Fig. 4C, co-treatment and HU were co-administered with 4NQO or B[a]P. In with MNU and O6-BG increased EGFP-MDC1 foci for- 4NQO-treated cells, co-treatment with AraC and HU mation at both 3 h and 24 h treatment but the increase signiˆcantly increased the EGFP-MDC1 foci area (Fig. was not signiˆcant. 4A).InthecaseofB[a])P, although signiˆcant EGFP- MDC1 foci formation requires 24 h of treatment (Fig. Discussion 3C and 4B), co-treatment with AraC and HU sig- Here, we established an easy-to-use genotoxicity as-

23 Shun Matsuda et al. say which combines EGFP-MDC1/MCF7 cells with an or DNA-PK but on ATR (74,75). After UV irradiation automated image-processing program using R software. or treatment with N-acetoxy-2-acetylamino‰uorene, This assay has several advantages compared with other which also forms bulky DNA adducts (76), g-H2AX in genotoxicity assays targeting g-H2AX. First, our G0/G1 phase is reduced in NER-defective cells such as method does not need any antibodies, which are expen- xerodrma pigmentosum (XP)-A, XP-C, or XP-G cells, sive consumables, whether obtained commercially or which are defective in formation of the pre-incision prepared in house. Second, our method does not require complex, recognition of DNA damage, and cleavage on any staining procedures and thus saves on staining and the 3? side of the damaged DNA, respectively handling time. Third, the assay accommodates live cell (74,75,77). Consistent with this, ectopic expression of imaging, which enables detection of not only EGFP- wild-type XPA in XP-A cells recovers the g-H2AX in-

MDC1 foci formation but also its detailed dynamic duction in G0/G1 phase after UV irradiation (74). Fur- changes in terms of location, size, shape and brightness. thermore, the g-H2AX in G0/G1 phase after UV irradia- CPT,4NQO,CDDP,andB[a]P are all known inducers tion is increased by DNA polymerase inhibition with of DSBs and g-H2AX (4,6,13,14,61–64), and all in- aphidicolin, AraC, or AraC/HU treatment, suggesting duced EGFP-MDC1 foci in our new genotoxicity assay that the DNA single-strand gaps formed during NER (Fig. 3). These results suggest that our assay can detect could eventually induce g-H2AX (74,75). We previously genotoxicity of a range of mutagens. utilized AraC/HU for sensitization of DNA strand CPT forms a stable complex with topoisomerase I break detection in a ‰uorometric analysis of DNA un- (Topo I) and DNA to inhibit Topo I (65) and induces a winding assay (78). These observations motivated us to DSB speciˆcally aŠecting the leading strand of the repli- test whether AraC/HU treatment would promote cation fork in S phase of the cell cycle (61). However, EGFP-MDC1 foci formation by bulky DNA adduct- we observed that EGFP-MDC1 foci formed in most forming mutagens which are repaired through NER. In cells treated with CPT (data not shown), suggesting that fact, AraC/HU treatment increased and accelerated CPT may also induce DSBs at stages other than S phase. EGFP-MDC1 foci formation in 4NQO-treated and Consistent with this possibility, it was reported that B[a]P-treated cells, respectively (Fig. 4), suggesting that CPT also forms a stable CPT-Topo I-DNA complex AraC/HU could sensitize our assay to mutagens during transcription (66) and induces DSBs and g- repaired through NER. Furthermore, our results strong- H2AX in post-mitotic cells (67). ly indicate that MDC1 participates not only in the cellu- 4NQO also induces DSBs (62) and the frequency is lar responses to DSBs and stalling of replication forks, not changed through the cell cycle (4). This fact is con- but also in NER-mediated repair. However, it is unclear sistent with our observation that EGFP-MDC1 foci whether g-H2AX/MDC1 foci respond to DNA single- were formed in most cells treated with 4NQO as well as strand gaps or the resultant DSB lesions. If g-H2AX/ CPT (data not shown). On the other hand, 4NQO most MDC1 foci directly respond to the former, then whether signiˆcantly induces g-H2AX foci in S phase (4), im- g-H2AX/MDC1 foci formation is a general process in plying that it generates DNA lesions other than DSBs NER will need to be elucidated in further studies. which also induce g-H2AX and MDC1 foci. One pos- MNU is a simple alkylating mutagen, and one of sibility is that stalled replication forks unaccompanied major DNA lesions induced by MNU is O6-MG (79). by DSBs result in foci formation. It was shown that MGMT repairs O6-MG to guanine by removing the bulky DNA adducts formed by 4NQO induce stalling of methyl adduct of the base (80,81). In this case, any replication forks in yeasts (68) and that stalling of repli- DNA single-strand gaps would not occur and thus there cation forks by mutagens induces ATR-dependent g- would be no opportunity for g-H2AX/MDC1 foci in- H2AX and MDC1 foci (16,41). However, at least to our duction as in NER. Our result, that MNU treatment did knowledge, it is still unclear what (DSBs, single-strand not increase EGFP-MDC1 foci (Fig. 3B), implies an im- breaks, or others?) actually induces g-H2AX and portant role for the MGMT repair system. Indeed, MDC1 foci at stalled replication forks. MGMT-defective cells induce g-H2AX in response to Bulky DNA adducts formed by mutagens such as MNU (10). These facts encouraged us that MGMT inhi- 4NQO, B[a]P, and ultraviolet (UV) irradiation are bition might sensitize our assay to MNU, and therefore repaired by nucleotide excision repair (NER) (69–72). we used the MGMT inhibitor O6-BG (82) for this pur- NER is a multistep process that consists of recognition pose. O6-BG binds to MGMT by competing with O6- of a DNA lesion, unwinding of the DNA helix around MG and transfers its benzyl group to MGMT, resulting the lesion, dual incision around the lesion, excision of in irreversible inhibition of MGMT (83). As shown in 24–32 nucleotides around the lesion, and gap-ˆlling Fig. 4C, co-treatment with MNU and O6-BG increased DNA synthesis (73). g-H2AX can also be induced in theareaofEGFP-MDC1focibuttheincreasewasnot cells in G0/G1 phase in an NER-dependent manner. In signiˆcant. It remains a possibility that optimization of this case, H2AX depends not on ATM O6-BG treatment conditions such as O6-BG concentra-

2424 Genotoxicity Assay Using EGFP-MDC1 tion and timing could improve the sensitivity of our as- interest. say to MNU. A major repair system of O6-MG other than MGMT is mismatch repair (MMR) (84,85). Since References MMR causes single-strand gaps through repair similar 1 Rogakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner to NER (86), gaps may induce g-H2AX through MMR WM. DNA double-stranded breaks induce histone H2AX as well as through NER. phosphorylation on serine 139. J Biol Chem. 1998; 273: As is well known, B[a]P requires metabolic activation 5858–68. for genotoxicity expression (87). In cells, B[a]P activates 2 Rogakou EP, Boon C, Redon C, Bonner WM. Megabase domains involved in DNA double-strand the transcription factor aryl hydrocarbon receptor breaks in vivo. J Cell Biol. 1999; 146: 905–16. which subsequently increases expression of xenobiotic 3 Rothkamm K, Lobrich M. Evidence for a lack of DNA metabolizing enzymes such as cytochrome P450 (CYP) double-strand break repair in human cells exposed to very 1A1 and CYP1A2 (88,89). These enzymes metabolize low x-ray doses. Proc Natl Acad Sci USA. 2003; 100: B[a]P to produce benzo(a)pyrene diolepoxide (BPDE) 5057–62. which forms a bulky DNA adduct (70,90). Thus, there is 4 Banath JP, Olive PL. Expression of phosphorylated hi- a time lag between enzyme induction and bulky DNA- stone H2AX as a surrogate of cell killing by drugs that adduct formation after B[a]P treatment (Supplemental create DNA double-strand breaks. Cancer Res. 2003; 63: Fig. 2). This is consistent with our data, that B[a]P 4347–50. needed 24 h to induce signiˆcant EGFP-MDC1 foci for- 5 Elmroth K, Nygren J, Martensson S, Ismail IH, Ham- mation whereas the direct mutagen 4NQO required only marsten O. Cleavage of cellular DNA by calicheamicin gamma1. DNA Repair (Amst). 2003; 2: 363–74. 1 h (Fig. 3). Furthermore, this result indicates that the 6 Liu JS, Kuo SR, Beerman TA, Melendy T. Induction of assay can utilize the bioactivation processes of the hu- DNA damage responses by adozelesin is S phase-speciˆc man cells and thus can detect genotoxicity of some in- and dependent on active replication forks. Mol Cancer direct mutagens without supplementary metabolic acti- Ther. 2003; 2: 41–7. vation although whether the cells su‹ciently express all 7 NoelG,GiocantiN,FernetM,Megnin-ChanetF, kinds of metabolic enzymes has not been tested. Favaudon V. Poly (ADP-ribose) polymerase (PARP-1) is Some mutagens such as UV can induce not only g- not involved in DNA double-strand break recovery. BMC H2AX foci but also pan-nuclear distribution of g- Cell Biol. 2003; 4: 7. H2AX (91). The image-processing program of our 8 Huang X, Traganos F, Darzynkiewicz Z. DNA damage genotoxicity assay may not detect the latter response as induced by DNA topoisomerase I- and topoisomerase II- positive and thus may underestimate the genotoxicity of inhibitors detected by histone H2AX phosphorylation in relation to the cell cycle phase and . Cell Cycle. some mutagens although we have not tested such muta- 2003; 2: 614–9. gens yet. Recent report showed that MDC1 binds to 9 HammondEM,DorieMJ,GiacciaAJ.ATR/ATMtar- ionizing irradiation-induced pan-nuclear g-H2AX, gets are phosphorylated by ATR in response to hypoxia clearly indicating involvement of MDC1 in DDR by and ATM in response to reoxygenation. J Biol Chem. pan-nuclear g-H2AX (92). A detailed observation of 2003; 278: 12207–13. cells to detect a change in localization of EGFP-MDC1 10 Komori K, Takagi Y, Sanada M, Lim TH, Nakatsu Y, between normal and pan-nuclear g-H2AX-forming cells Tsuzuki T, et al. A novel protein, MAPO1, that functions would be required in a further study for improvement of in apoptosis triggered by O6-methylguanine mispair in our genotoxicity assay. DNA. . 2009; 28: 1142–50. The protocol of our assay is simple (mutagen treat- 11 Stojic L, Mojas N, Cejka P, Di Pietro M, Ferrari S, ment, microscopic imaging, and image-processing and Marra G, et al. Mismatch repair-dependent G2 check- point induced by low doses of SN1 type methylating calculation of foci area/nucleus using R). Therefore, we agents requires the ATR kinase. Genes Dev. 2004; 18: are conˆdent that automation of all the steps can be 1331–44. achieved and thus our assay can be applied to high- 12 Liu JS, Kuo SR, Melendy T. Comparison of checkpoint throughput genotoxicity screening and monitoring of responses triggered by DNA polymerase inhibition versus environmental mutagens. DNA damaging agents. Mutat Res. 2003; 532: 215–26. 13 Zhou C, Li Z, Diao H, Yu Y, Zhu W, Dai Y, et al. DNA Acknowledgements: This study was supported by damage evaluated by gammaH2AX foci formation by a KAKENHI (23221006) from JSPS, Grant-in-Aid for selective group of chemical/physical stressors. Mutat JSPS Fellows (24–5844), and Grant-in-Aid for Scientiˆc Res. 2006; 604: 8–18. Research on Innovative Areas ``Coupling of replication, 14 Huang X, Okafuji M, Traganos F, Luther E, Holden E, repair and transcription, and their common mechanism Darzynkiewicz Z. Assessment of histone H2AX phos- phorylation induced by DNA topoisomerase I and II inhi- of '' from JSPS. bitors topotecan and mitoxantrone and by the DNA cross-linking agent cisplatin. Cytometry A. 2004; 58: Con‰icts of interest: The authors declare no con‰ict of

2525 Shun Matsuda et al.

99–110. icity of nano-SiO2 on human epithelial intestinal HT-29 15 Mischo HE, Hemmerich P, Grosse F, Zhang S. Ac- cell line. Ann Occup Hyg. 2012; 56: 622–30. tinomycin D induces histone gamma-H2AX foci and 31 WanR,MoY,FengL,ChienS,TollerudDJ,ZhangQ. complex formation of gamma-H2AX with Ku70 and DNA damage caused by metal nanoparticles: involve- nuclear DNA helicase II. J Biol Chem. 2005; 280: ment of oxidative stress and activation of ATM. Chem 9586–94. Res Toxicol. 2012; 25: 1402–11. 16 Ward IM, Chen J. Histone H2AX is phosphorylated in 32 Toduka Y, Toyooka T, Ibuki Y. Flow cytometric evalua- an ATR-dependent manner in response to replicational tion of nanoparticles using side-scattered light and reac- stress. J Biol Chem. 2001; 276: 47759–62. tive oxygen species-mediated ‰uorescence-correlation 17 Hammond EM, Green SL, Giaccia AJ. Comparison of with genotoxicity. Environ Sci Technol. 2012; 46: hypoxia-induced replication arrest with hydroxyurea and 7629–36. aphidicolin-induced arrest. Mutat Res. 2003; 532: 33 OzakiT,NagaseT,IchimiyaS,SekiN,OhiriM,Nomura 205–13. N, et al. NFBD1/KIAA0170 is a novel nuclear transcrip- 18 Andreau K, Castedo M, Perfettini JL, Roumier T, tional transactivator with BRCT domain. DNA Cell Biol. Pichart E, Souquere S, et al. Preapoptotic chromatin 2000; 19: 475–85. condensation upstream of the mitochondrial checkpoint. 34 de Jager M, van Noort J, van Gent DC, Dekker C, J Biol Chem. 2004; 279: 55937–45. Kanaar R, Wyman C. Human Rad50/Mre11 is a ‰exible 19 Toyooka T, Ishihama M, Ibuki Y. Phosphorylation of hi- complex that can tether DNA ends. Mol Cell. 2001; 8: stone H2AX is a powerful tool for detecting chemical 1129–35. photogenotoxicity. J Invest Dermatol. 2011; 131: 35 You Z, Chahwan C, Bailis J, Hunter T, Russell P. ATM 1313–21. activation and its recruitment to damaged DNA require 20 Toyooka T, Ibuki Y. Co-exposure to benzo[a]pyrene and binding to the C terminus of Nbs1. Mol Cell Biol. 2005; UVA induces phosphorylation of histone H2AX. FEBS 25: 5363–79. Lett. 2005; 579: 6338–42. 36 Falck J, Coates J, Jackson SP. Conserved modes of 21 Ishihama M, Toyooka T, Ibuki Y. Generation of phos- recruitment of ATM, ATR and DNA-PKcs to sites of phorylated histone H2AX by benzene metabolites. Tox- DNA damage. Nature. 2005; 434: 605–11. icol In Vitro. 2008; 22: 1861–8. 37 Burma S, Chen BP, Murphy M, Kurimasa A, Chen DJ. 22 Shimohara C, Sawai T, Yagi T. Polyaromatic hydrocar- ATM phosphorylates histone H2AX in response to DNA bons cause histone H2AX phosphorylation in the S-phase double-strand breaks. J Biol Chem. 2001; 276: 42462–7. of the cell cycle. Genes Environ. 2008; 30: 92–9. 38 StiŠ T, O'Driscoll M, Rief N, Iwabuchi K, Lobrich M, 23 Toyooka T, Kubota T, Ibuki Y. Nonylphenol polyethox- Jeggo PA. ATM and DNA-PK function redundantly to ylates induce phosphorylation of histone H2AX. Mutat phosphorylate H2AX after exposure to ionizing radia- Res. 2012; 741: 57–64. tion. Cancer Res. 2004; 64: 2390–6. 24 Ha L, Ceryak S, Patierno SR. Generation of S phase- 39 Stucki M, Clapperton JA, Mohammad D, YaŠe MB, dependent DNA double-strand breaks by Cr (VI) ex- Smerdon SJ, Jackson SP. MDC1 directly binds phos- posure: involvement of ATM in Cr (VI) induction of phorylated histone H2AX to regulate cellular responses gamma-H2AX. Carcinogenesis. 2004; 25: 2265–74. to DNA double-strand breaks. Cell. 2005; 123: 1213–26. 25 Albino AP, Jorgensen ED, Rainey P, Gillman G, Clark 40 Lou Z, Minter-Dykhouse K, Franco S, Gostissa M, TJ, Gietl D, et al. gammaH2AX: A potential DNA Rivera MA, Celeste A, et al. MDC1 maintains genomic damage response biomarker for assessing toxicological stability by participating in the ampliˆcation of ATM- risk of tobacco products. Mutat Res. 2009; 678: 43–52. dependent DNA damage signals. Mol Cell. 2006; 21: 26 Pacurari M, Yin XJ, Zhao J, Ding M, Leonard SS, 187–200. Schwegler-Berry D, et al. Raw single-wall carbon 41 Wang J, Gong Z, Chen J. MDC1 collaborates with nanotubes induce oxidative stress and activate MAPKs, TopBP1 in DNA replication checkpoint control. J Cell AP-1, NF-kappaB, and Akt in normal and malignant hu- Biol. 2011; 193: 267–73. man mesothelial cells. Environ Health Perspect. 2008; 42 Shang YL, Bodero AJ, Chen PL. NFBD1, a novel 116: 1211–7. nuclear protein with signature motifs of FHA and BRCT, 27 Ahamed M, Karns M, Goodson M, Rowe J, Hussain SM, and an internal 41- repeat sequence, is an early Schlager JJ, et al. DNA damage response to diŠerent sur- participant in DNA damage response. J Biol Chem. 2003; face chemistry of silver nanoparticles in mammalian cells. 278: 6323–9. Toxicol Appl Pharmacol. 2008; 233: 404–10. 43 So S, Davis AJ, Chen DJ. Autophosphorylation at serine 28 Trouiller B, Reliene R, Westbrook A, Solaimani P, 1981 stabilizes ATM at DNA damage sites. J Cell Biol. Schiestl RH. Titanium dioxide nanoparticles induce DNA 2009; 187: 977–90. damage and genetic instability in vivo in mice. Cancer 44 Lou Z, Minter-Dykhouse K, Wu X, Chen J. MDC1 is Res. 2009; 69: 8784–9. coupled to activated CHK2 in mammalian DNA damage 29 Tsaousi A, Jones E, Case CP. The in vitro genotoxicity of response pathways. Nature. 2003; 421: 957–61. orthopaedic ceramic (Al2O3) and metal (CoCr alloy) par- 45 Zhang J, Ma Z, Treszezamsky A, Powell SN. MDC1 in- ticles. Mutat Res. 2010; 697: 1–9. teracts with Rad51 and facilitates homologous recombi- 30 Sergent JA, Paget V, Chevillard S. Toxicity and genotox- nation. Nat Struct Mol Biol. 2005; 12: 902–9.

2626 Genotoxicity Assay Using EGFP-MDC1

46 Wu L, Luo K, Lou Z, Chen J. MDC1 regulates intra-S- BImage—an R package for image processing with appli- phase checkpoint by targeting NBS1 to DNA double- cations to cellular phenotypes. Bioinformatics. 2010; 26: strand breaks. Proc Natl Acad Sci USA. 2008; 105: 979–81. 11200–5. 61 Strumberg D, Pilon AA, Smith M, Hickey R, Malkas L, 47 Spycher C, Miller ES, Townsend K, Pavic L, Morrice Pommier Y. Conversion of topoisomerase I cleavage NA, Janscak P, et al. Constitutive phosphorylation of complexes on the leading strand of ribosomal DNA into MDC1 physically links the MRE11-RAD50-NBS1 com- 5?-phosphorylated DNA double-strand breaks by replica- plex to damaged chromatin. J Cell Biol. 2008; 181: tion runoŠ. Mol Cell Biol. 2000; 20: 3977–87. 227–40. 62 Andoh T, Ide T. Strand scission and rejoining of DNA in 48 Melander F, Bekker-Jensen S, Falck J, Bartek J, cultured mammalian cells induced by 4-nitroquinoline 1- Mailand, N, Lukas, J. Phosphorylation of SDT repeats oxide. Cancer Res. 1972; 32: 1230–5. in the MDC1 N terminus triggers retention of NBS1 at the 63 Frankenberg-Schwager M, Kirchermeier D, Greif G, Baer DNA damage-modiˆed chromatin. J Cell Biol. 2008; 181: K, Becker M, Frankenberg D. Cisplatin-mediated DNA 213–26. double-strand breaks in replicating but not in quiescent 49 Chapman JR, Jackson SP. Phospho-dependent interac- cells of the yeast Saccharomyces cerevisiae. Toxicology. tions between NBS1 and MDC1 mediate chromatin reten- 2005; 212: 175–84. tion of the MRN complex at sites of DNA damage. 64 Celotti L, Ferraro P, Furlan D, Zanesi N, Pavanello S. EMBO Rep. 2008; 9: 795–801. DNA repair in human lymphocytes treated in vitro with 50 Mailand N, Bekker-Jensen S, Faustrup H, Melander F, (＋)-anti- and (＋/-)-syn-benzo[a]pyrene diolepoxide. Bartek J, Lukas C, et al. RNF8 ubiquitylates at Mutat Res. 1993; 294: 117–26. DNA double-strand breaks and promotes assembly of 65 HsiangYH,HertzbergR,HechtS,LiuLF.Camptothe- repair proteins. Cell. 2007; 131: 887–900. cin induces protein-linked DNA breaks via mammalian 51 HuenMS,GrantR,MankeI,MinnK,YuX,YaŠeMB, DNA topoisomerase I. J Biol Chem. 1985; 260: 14873–8. et al. RNF8 transduces the DNA-damage signal via hi- 66 Bendixen C, Thomsen B, Alsner J, Westergaard O. Cam- stone ubiquitylation and checkpoint protein assembly. ptothecin-stabilized topoisomerase I-DNA adducts cause Cell. 2007; 131: 901–14. premature termination of transcription. Biochemistry. 52 Kolas NK, Chapman JR, Nakada S, Ylanko J, Chahwan 1990; 29: 5613–9. R, Sweeney FD, et al. Orchestration of the DNA-damage 67 Sordet O, Redon CE, Guirouilh-Barbat J, Smith S, Solier response by the RNF8 . Science. 2007; S, Douarre C, et al. Ataxia telangiectasia mutated activa- 318: 1637–40. tion by transcription- and topoisomerase I-induced DNA 53 Lou Z, Chen BP, Asaithamby A, Minter-Dykhouse K, double-strand breaks. EMBO Rep. 2009; 10: 887–93. Chen DJ, Chen J. MDC1 regulates DNA-PK au- 68 Minca EC, Kowalski D. Replication fork stalling by tophosphorylation in response to DNA damage. J Biol bulky DNA damage: localization at active origins and Chem. 2004; 279: 46359–62. checkpoint modulation. Nucleic Acids Res. 2011; 39: 54 Nakanishi M, Ozaki T, Yamamoto H, Hanamoto T, 2610–23. KikuchiH,FuruyaK,etal.NFBD1/MDC1associates 69 Jones CJ, Edwards SM, Waters R. The repair of identi- with p53 and regulates its function at the crossroad be- ˆed large DNA adducts induced by 4-nitroquinoline-1- tween cell survival and death in response to DNA oxide in normal or xeroderma pigmentosum group A hu- damage. J Biol Chem. 2007; 282: 22993–3004. man ˆbroblasts, and the role of DNA polymerases alpha 55 Eliezer Y, Argaman L, Rhie A, Doherty AJ, Goldberg M. or delta. Carcinogenesis. 1989; 10: 1197–201. The direct interaction between 53BP1 and MDC1 is re- 70 Pulkrabek P, LeŒer S, Grunberger D, Weinstein IB. quired for the recruitment of 53BP1 to sites of damage. J Modiˆcation of deoxyribonucleic acid by a diol epoxide Biol Chem. 2009; 284: 426–35. of benzo[a]pyrene. Relation to deoxyribonucleic acid 56 Coster G, Hayouka Z, Argaman L, Strauss C, Friedler A, structure and conformation and eŠects on transfectional Brandeis M, et al. The DNA damage response mediator activity. Biochemistry. 1979; 18: 5128–34. MDC1 directly interacts with the anaphase-promoting 71 Cleaver JE. Defective repair replication of DNA in complex/cyclosome. J Biol Chem. 2007; 282: 32053–64. xeroderma pigmentosum. Nature. 1968; 218: 652–6. 57 Inoue K, Nakanjishi M, Kikuchi H, Yamamoto H, Todo 72 SetlowRB,ReganJD,GermanJ,CarrierWL.Evidence S, Nakagawara A, et al. NFBD1/MDC1 stabilizes onco- that xeroderma pigmentosum cells do not perform the genic MDM2 to contribute to cell fate determination in ˆrst step in the repair of ultraviolet damage to their DNA. response to DNA damage. Biochem Biophys Res Com- Proc Natl Acad Sci USA. 1969; 64: 1035–41. mun. 2008; 371: 829–33. 73 VolkerM,MoneMJ,KarmakarP,vanHoŠenA,Schul 58 Luo K, Yuan J, Chen J, Lou Z. Topoisomerase IIalpha W, Vermeulen W, et al. Sequential assembly of the controls the decatenation checkpoint. Nat Cell Biol. nucleotide excision repair factors in vivo. Mol Cell. 2001; 2009; 11: 204–10. 8: 213–24. 59 Ihaka R, Gentleman R. R: A language for data analysis 74 Matsumoto M, Yaginuma K, Igarashi A, Imura M, and graphics. Journal of computational and graphical Hasegawa M, Iwabuchi K, et al. Perturbed gap-ˆlling statistics. 1996; 3: 299–314. synthesis in nucleotide excision repair causes histone 60 Pau G, Fuchs F, Sklyar O, Boutros M, Huber W. E- H2AX phosphorylation in human quiescent cells. J Cell

2727 Shun Matsuda et al.

Sci. 2007; 120: 1104–12. nine. Biochemistry. 1993; 32: 11998–2006. 75 Hanasoge S, Ljungman M. H2AX phosphorylation after 84 Ceccotti S, Aquilina G, Macpherson P, Yamada M, Kar- UV irradiation is triggered by DNA repair intermediates ran P, Bignami M. Processing of O6-methylguanine by and is mediated by the ATR kinase. Carcinogenesis. mismatch correction in human cell extracts. Curr Biol. 2007; 28: 2298–304. 1996; 6: 1528–31. 76 Vu VT, Grantham PH, Roller PP, Hankins WD, Wirth 85 Duckett DR, Drummond JT, Murchie AI, Reardon JT, PJ, Thorgeirsson SS. Formation of DNA adducts from Sancar A, Lilley DM, et al. Human MutSalpha recog- N-acetoxy-2-acetylamino‰uorene and N-hydroxy-2- nizes damaged DNA base pairs containing O6-methylgua- acetylamino‰uorene in rat hemopoietic tissues in vivo. nine, O4-methylthymine, or the cisplatin-d (GpG) ad- Cancer Res. 1986; 46: 233–8. duct. Proc Natl Acad Sci USA. 1996; 93: 6443–7. 77 Marti TM, Hefner E, Feeney L, Natale V, Cleaver JE. 86 Grilley M, Gri‹th J, Modrich P. Bidirectional excision in H2AX phosphorylation within the G1 phase after UV ir- methyl-directed mismatch repair. J Biol Chem. 1993; 268: radiation depends on nucleotide excision repair and not 11830–7. DNA double-strand breaks. Proc Natl Acad Sci USA. 87 Shimizu Y, Nakatsuru Y, Ichinose M, Takahashi Y, 2006; 103: 9891–6. Kume H, Mimura J, et al. Benzo[a]pyrene carcinogenici- 78 MatsudaT,YagiT,TakebeH,MatsuiS.Detectionand ty is lost in mice lacking the aryl hydrocarbon receptor. quantiˆcation of DNA strand breaks in human cells in- Proc Natl Acad Sci USA. 2000; 97: 779–82. duced by contaminants in Japanese tap waters. Water 88 Harper PA, Prokipcak RD, Bush LE, Golas CL, Okey Science and Technology. 1996; 33: 297–304. AB. Detection and characterization of the Ah receptor 79 Pegg AE. Methylation of the O6 position of guanine in for 2,3,7,8-tetrachlorodibenzo-p-dioxin in the human col- DNA is the most likely initiating event in carcinogenesis on adenocarcinoma cell line LS180. Arch Biochem by methylating agents. Cancer Invest. 1984; 2: 223–31. Biophys. 1991; 290: 27–36. 80 Zak P, Kleibl K, Laval F. Repair of O6-methylguanine 89 Nebert DW, Roe AL, Dieter MZ, Solis WA, Yang Y, and O4-methylthymine by the human and rat O6-methyl- Dalton TP. Role of the aromatic hydrocarbon receptor guanine-DNA methyltransferases. J Biol Chem. 1994; and [Ah] battery in the oxidative stress response, cell 269: 730–3. cycle control, and apoptosis. Biochem Pharmacol. 2000; 81 Oh HK, Teo AK, Ali RB, Lim A, Ayi TC, Yarosh DB, et 59: 65–85. al. Conformational change in human DNA repair enzyme 90 Gelboin HV. Benzo[alpha]pyrene metabolism, activation O6-methylguanine-DNA methyltransferase upon alkyla- and carcinogenesis: role and regulation of mixed-function tion of its active site by SN1 (indirect-acting) and SN2 oxidases and related enzymes. Physiol Rev. 1980; 60: (direct-acting) alkylating agents: breaking a ``salt-link''. 1107–66. Biochemistry. 1996; 35: 12259–66. 91 de Feraudy S, Revet I, Bezrookove V, Feeney L, Cleaver 82 Dolan ME, Moschel RC, Pegg AE. Depletion of mam- JE. A minority of foci or pan-nuclear apoptotic staining malian O6-alkylguanine-DNA alkyltransferase activity by of gammaH2AX in the S phase after UV damage contain O6-benzylguanine provides a means to evaluate the role DNA double-strand breaks. Proc Natl Acad Sci USA. of this protein in protection against carcinogenic and 2010; 107: 6870–5. therapeutic alkylating agents. Proc Natl Acad Sci USA. 92 Meyer B, Voss KO, Tobias F, Jakob B, Durante M, 1990; 87: 5368–72. Taucher-Scholz G. Clustered DNA damage induces pan- 83 PeggAE,BoosalisM,SamsonL,MoschelRC,ByersTL, nuclear H2AX phosphorylation mediated by ATM and Swenn K, et al. Mechanism of inactivation of human O6- DNA-PK. Nucleic Acids Res. 2013; 41: 6109–18. alkylguanine-DNA alkyltransferase by O6-benzylgua-

2828

Top View